ReleaseEngineering/Meetings/2007-jun-18

From FedoraProject

Jump to: navigation, search

Contents

Fedora Release Engineering Meeting :: Monday 2007-06-18

Schedule

  • 20-July is the freeze date for Test1
  • Prepare for Test1 starting on 13-July

Decisions:

  • Release Engineering will deliver the following:

1. Fedora spin which is iso(s) + exploded installable tree 1. Fedora Live images for i386/x86_64 that are installable. 1. KDE Live images that are installable 1. Everything repo of packages that we can hopefully use with other install media to "make" it installable.

  • Various SIGs may be responsible for content definition for these various deliverables, as well as pre-generation and QA of them.
  • Release Engineering is only responsible for final production and signing of the content.
  • At least one week before Test1 freeze (13-July) rel-eng will be looking closer and attempting to fix up issues such as:

1. broken deps 1. broken upgrade paths 1. conflicts 1. other such tree issues. 1. Compose tools

  • should be somewhat stabilized (no changes to code) before this point so that pre-freeze trees can be generated and reviewed before the mad freeze rush.

Open Discussion

IRC Transcript

Idea.png f13 changed the topic of #fedora-meeting to: Fedora Release Engineering Meeting<a href="#t13:00" class="time">13:00</a>
jwbwhee<a href="#t13:01" class="time">13:01</a>
* jeremy is here<a href="#t13:01" class="time">13:01</a>
f13rdieter: warren: jwb: poelcat: wwoods: ping<a href="#t13:01" class="time">13:01</a>
* lmacken is here for a few<a href="#t13:01" class="time">13:01</a>
jwbi said whee man<a href="#t13:01" class="time">13:01</a>
jwb;)<a href="#t13:01" class="time">13:01</a>
* rdieter is about 1/2 here.<a href="#t13:01" class="time">13:01</a>
f13jwb: gah, sorry.  blame Monday<a href="#t13:02" class="time">13:02</a>
jwbindeed<a href="#t13:02" class="time">13:02</a>
* mmcgrath pongs<a href="#t13:02" class="time">13:02</a>
* warren here<a href="#t13:02" class="time">13:02</a>
warrenalthough i'm eating<a href="#t13:03" class="time">13:03</a>
f13so...<a href="#t13:03" class="time">13:03</a>
f13we don't actually have any topics listed.<a href="#t13:03" class="time">13:03</a>
warrendid we lock into the release schedule?<a href="#t13:04" class="time">13:04</a>
warren(especially for test1)<a href="#t13:04" class="time">13:04</a>
f13warren: rel-eng didn't necessarily.<a href="#t13:04" class="time">13:04</a>
Idea.png f13 changed the topic of #fedora-meeting to: Fedora Release Engineering Meeting - Schedule<a href="#t13:04" class="time">13:04</a>
* wwoods here<a href="#t13:04" class="time">13:04</a>
f13so July 20th is the freeze date for TEst1<a href="#t13:05" class="time">13:05</a>
poelcathere<a href="#t13:05" class="time">13:05</a>
* abadger1999 here<a href="#t13:05" class="time">13:05</a>
f13I'd like to have pungi in a usable state a week or so before that, spinning some test trees<a href="#t13:05" class="time">13:05</a>
f13and not making changes to the compose software leading up to the freeze<a href="#t13:05" class="time">13:05</a>
f13the week before would be a good time to start looking at broken deps, broken upgrade paths, etc...<a href="#t13:05" class="time">13:05</a>
f13things that could easily be fixed duirng that week so that the freeze isn't a mess.<a href="#t13:06" class="time">13:06</a>
* bpepple_laptop lurks in the background.<a href="#t13:07" class="time">13:07</a>
f13any other thoughts about how to prepare for the test release?<a href="#t13:07" class="time">13:07</a>
jwbcan't we hack mash to puke on broken deps?<a href="#t13:07" class="time">13:07</a>
f13I imagine I'll probably be doing the composes on my workstations again.<a href="#t13:07" class="time">13:07</a>
f13jwb: we don't want it to puke, just complain loudly<a href="#t13:08" class="time">13:08</a>
f13pungi complains about broken deps too<a href="#t13:08" class="time">13:08</a>
jwbok, but i mean couldn't we start that now?<a href="#t13:08" class="time">13:08</a>
mclasenare delta rpms going to be supported for fc8 ?<a href="#t13:08" class="time">13:08</a>
jeremyf13: I'd really prefer to not be doing the composes on workstations<a href="#t13:08" class="time">13:08</a>
jwbso that around test time, it should be a small number<a href="#t13:08" class="time">13:08</a>
jeremyf13: (both for the pungi trees and the live images)<a href="#t13:08" class="time">13:08</a>
jeremymclasen: "maybe"<a href="#t13:08" class="time">13:08</a>
f13jeremy: well, mock chroot here vs mock chroot in the colo doesn't make al ot of difference to me, other than the time it takes to actually test the tree once composed.<a href="#t13:09" class="time">13:09</a>
jeremymclasen: I see jdieter sent mail to fedora-devel-list and I was hoping to get a mail out to start some discussion... so we'll have to see where it goes and what can be done<a href="#t13:09" class="time">13:09</a>
f13jwb: the broken dep report goes out every night at the bottom of rawhide report.  The question is poking maintainers to fix their crap<a href="#t13:09" class="time">13:09</a>
jeremyf13: if we're going to get beyond it being you and me testing the tree, it makes a huge difference<a href="#t13:09" class="time">13:09</a>
f13jeremy: sure<a href="#t13:09" class="time">13:09</a>
jeremy(and wwoods and other folks inside the fence so to speak)<a href="#t13:09" class="time">13:09</a>
f13need space (:<a href="#t13:10" class="time">13:10</a>
jeremyand then the kde folks can do their live images themselves, etc, etc<a href="#t13:10" class="time">13:10</a>
f13so I guess there is one other thing to talk about, what are we going to produce for Test1.<a href="#t13:10" class="time">13:10</a>
f13a Fedora spin,  Live images, KDE Live images, do we want to target an installable Everything tree?<a href="#t13:10" class="time">13:10</a>
f13probably should if we can<a href="#t13:11" class="time">13:11</a>
wwoodsmight as well drop some images in there, I guess<a href="#t13:11" class="time">13:11</a>
wwoodsbut.. I mean, why even bother with multiple trees in that case?<a href="#t13:11" class="time">13:11</a>
mclasenwas there talk about doing weekly (or so) live cd spins earlier ?<a href="#t13:11" class="time">13:11</a>
wwoodsKDE/Fedora/Live images and the installable Everything tree<a href="#t13:12" class="time">13:12</a>
jeremyf13: I'd rather target getting the installer able to point at the everything tree rather than an installable everything tree<a href="#t13:12" class="time">13:12</a>
jeremymclasen: there had been, yes.  and I've tried each of the past two weeks but not had it fit on a cd<a href="#t13:12" class="time">13:12</a>
jeremymclasen: I kicked one off this morning, though, and it seems to fit now.  so I was actually going to put it up later after making sure it boots<a href="#t13:12" class="time">13:12</a>
f13jeremy: ok.<a href="#t13:13" class="time">13:13</a>
rdieterwwoods: +1, simpler, minimize user confusion about which iso/spin to use.<a href="#t13:13" class="time">13:13</a>
f13wwoods: because downloading 3 DVD images to be able to do a non-network install is rather silly<a href="#t13:13" class="time">13:13</a>
wwoods3 DVD images? who said anything about doing Everything media?<a href="#t13:13" class="time">13:13</a>
f13wwoods: gotta have media for upgrades and non-network traditional installs<a href="#t13:14" class="time">13:14</a>
f13unless we want to play the 'you get what we want you to have whether you like it or not' game.<a href="#t13:14" class="time">13:14</a>
* rdieter forgot about non-network upgrades...<a href="#t13:14" class="time">13:14</a>
wwoodsright, which is what the Fedora media is for<a href="#t13:14" class="time">13:14</a>
f13wwoods: maybe we're talking past eachother.<a href="#t13:15" class="time">13:15</a>
wwoodscurrently we offer: Fedora media, Fedora tree, Live media, Everything tree<a href="#t13:15" class="time">13:15</a>
wwoodsEverything tree has no installer images in it<a href="#t13:15" class="time">13:15</a>
wwoodsFedora tree does<a href="#t13:15" class="time">13:15</a>
f13and your proposal?<a href="#t13:16" class="time">13:16</a>
wwoodsdrop the Fedora tree, get installer images in the Everything tree<a href="#t13:16" class="time">13:16</a>
wwoodskeep the media as-is<a href="#t13:16" class="time">13:16</a>
warrenspin != tree...<a href="#t13:16" class="time">13:16</a>
warrenI wonder if that will confuse anyone<a href="#t13:17" class="time">13:17</a>
f13they're already confused :(<a href="#t13:17" class="time">13:17</a>
wwoodsI'm aware of the difference. You'll notice that I haven't mentioned spins at all.<a href="#t13:17" class="time">13:17</a>
wwoodsyep, they are. I'm talking only in terms of Things We Deliver To Users<a href="#t13:17" class="time">13:17</a>
warrenf13, true<a href="#t13:17" class="time">13:17</a>
* jeremy raises his hand as confused<a href="#t13:17" class="time">13:17</a>
f13heh<a href="#t13:17" class="time">13:17</a>
jwbditto<a href="#t13:17" class="time">13:17</a>
f13wwoods: it'll make it harder to have a mirror of the exploaded Fedora tree<a href="#t13:17" class="time">13:17</a>
f13but not impossible.<a href="#t13:17" class="time">13:17</a>
wwoodswhy do we want exploded Fedora trees at all?<a href="#t13:18" class="time">13:18</a>
wwoodsif someone's gonna do a network install why wouldn't they install from the Everything tree instead<a href="#t13:18" class="time">13:18</a>
nottingwwoods: smaller thing to mirror for ftp/http installs?<a href="#t13:18" class="time">13:18</a>
nottingalso, the metdata is different<a href="#t13:18" class="time">13:18</a>
wwoodssince it's a superset of Fedora you lose nothing<a href="#t13:18" class="time">13:18</a>
wwoodsand if you want a local mirror of Fedora you can loop-mount the iso<a href="#t13:18" class="time">13:18</a>
f13wwoods: that doesn't work<a href="#t13:18" class="time">13:18</a>
f13media:// urls in iso repodata<a href="#t13:19" class="time">13:19</a>
wwoodsbunk!<a href="#t13:19" class="time">13:19</a>
f13and I think the anaconda team just did away with loopback moutned isos install support<a href="#t13:19" class="time">13:19</a>
* warren still waiting for anaconda that can install from an ISO over http.<a href="#t13:19" class="time">13:19</a>
f13rather ftp served loopback mounted isos<a href="#t13:19" class="time">13:19</a>
wwoodsloopback mounted iso *sets*<a href="#t13:19" class="time">13:19</a>
jeremyf13: media:// urls should be fine<a href="#t13:19" class="time">13:19</a>
jeremyfwiw<a href="#t13:19" class="time">13:19</a>
* notting missed the context. what are we talking about?<a href="#t13:19" class="time">13:19</a>
f13jeremy: as of when?  Lots of people are having trouble with the Fedora 7 isos<a href="#t13:20" class="time">13:20</a>
jeremywwoods: metadata for everything tree is substantially larger. ==> higher memory usage<a href="#t13:20" class="time">13:20</a>
jeremyf13: yeah, it broke.  but fixed up for yum 3.2.1<a href="#t13:20" class="time">13:20</a>
wwoodsjeremy: true<a href="#t13:20" class="time">13:20</a>
f13notting: wwoods is proposing dropping the exploaded "Fedora" spin tree.<a href="#t13:20" class="time">13:20</a>
f13notting: having /only/ the Everything exploaded tree<a href="#t13:20" class="time">13:20</a>
wwoodsI'm proposing we have exactly one installable exploded tree on the mirrors, rather than two exploded trees<a href="#t13:20" class="time">13:20</a>
wwoodsone of which is installable (with boot images)<a href="#t13:20" class="time">13:20</a>
f13wwoods: is the crux of the problem you're trying to fix is confusion as to which tree to mirror?<a href="#t13:20" class="time">13:20</a>
wwoodsthe other one.. might be installable? no images though<a href="#t13:20" class="time">13:20</a>
wwoodsthe crux of the problem is that there's two exploded trees and no clear answer about which one people should use to install / mirror<a href="#t13:21" class="time">13:21</a>
tibbsI have to say I found it very strange that Everything doesn't even have the PXE boot images.<a href="#t13:22" class="time">13:22</a>
* rdieter nods, agrees that wwoods has a point.<a href="#t13:22" class="time">13:22</a>
f13wwoods: but isn't that really boil down to "there are multiple spins of Fedora" ?<a href="#t13:22" class="time">13:22</a>
f13because the isos not matching the tree, that's not going to be confusing at all.<a href="#t13:22" class="time">13:22</a>
warrenIs our official recommendation for folks who want a single disc installer to use live?<a href="#t13:22" class="time">13:22</a>
wwoodswarren: single-CD, you mean?<a href="#t13:22" class="time">13:22</a>
warren(Is it possible to install from live without graphics?)<a href="#t13:23" class="time">13:23</a>
f13a dvd is a single disc...<a href="#t13:23" class="time">13:23</a>
warrenyes, single-CD text-mode install possible?<a href="#t13:23" class="time">13:23</a>
wwoodsthe ISOs have never matched the full tree (core+extras)<a href="#t13:23" class="time">13:23</a>
f13wwoods: Extras was never in the tree though, it was always in a different directory<a href="#t13:23" class="time">13:23</a>
wwoodsright, in the past there was exactly one installable tree and one set of media<a href="#t13:24" class="time">13:24</a>
f13wwoods: I'm not saying no, I'm just saying it may trade one set of confusion for another, due to the basic fact of having multiple spins is confusing.<a href="#t13:24" class="time">13:24</a>
nottingand it isn't really going to get *less* confusing<a href="#t13:25" class="time">13:25</a>
wwoodsI guess? I feel like having one installable tree is simpler than having two<a href="#t13:25" class="time">13:25</a>
jeremynotting: indeed :-/<a href="#t13:25" class="time">13:25</a>
wwoodsactually, hah<a href="#t13:26" class="time">13:26</a>
wwoodswe still have one installable tree<a href="#t13:26" class="time">13:26</a>
f13yeah, but that was more because we ran out of runway<a href="#t13:26" class="time">13:26</a>
wwoodsbut right now there's one installable and one mysteriously uninstallable tree<a href="#t13:26" class="time">13:26</a>
f13rather than by design.<a href="#t13:26" class="time">13:26</a>
wwoodsso, you know, it's cool if we have two installable trees, one of which is a subset of the other<a href="#t13:26" class="time">13:26</a>
wwoodsit might be simpler just to have one big installable tree but I don't really care<a href="#t13:26" class="time">13:26</a>
wwoodsso long as they're both installable I'm happier<a href="#t13:26" class="time">13:26</a>
f13also the layout was created with the idea that there would be multiple subset spins, not just "Fedora"<a href="#t13:26" class="time">13:26</a>
warrenone installable tree might be easier to test for internal consistency<a href="#t13:27" class="time">13:27</a>
f13warren: not going to get away with that.<a href="#t13:27" class="time">13:27</a>
nottingbut the Future is multiple trees. for better or worse<a href="#t13:27" class="time">13:27</a>
* jwb has phone call<a href="#t13:28" class="time">13:28</a>
f13jeremy: can you expand a bit upon: <jeremy> f13: I'd really prefer to not be doing the composes on workstations<a href="#t13:28" class="time">13:28</a>
f13er<a href="#t13:28" class="time">13:28</a>
f13<jeremy> f13: I'd rather target getting the installer able to point at the everything tree rather than an installable everything tree<a href="#t13:28" class="time">13:28</a>
nottingjeremy: you want the installer to be able to select stuff for everything? i can commit that patch really fast :)<a href="#t13:29" class="time">13:29</a>
jeremyf13: basically, making the change so that rescuecd.iso --> installcd.iso.  and then being able to add the repo in the second stage rather than having to do it in the loader<a href="#t13:29" class="time">13:29</a>
jeremynotting: that'd be good too, but is sort of separate<a href="#t13:30" class="time">13:30</a>
f13jeremy: ok, so the Everything tree would just be repodata adn the installer wouldn't care?<a href="#t13:31" class="time">13:31</a>
f13(basically what we have today)<a href="#t13:31" class="time">13:31</a>
jeremyf13: yes, but making things a little cleaner<a href="#t13:32" class="time">13:32</a>
nottingjeremy: so, you just want rescuecd to be a single-stage (hee) installer that can be pointed at an arbitrary tree?<a href="#t13:32" class="time">13:32</a>
jeremynotting: *thwap* ;-)    there'd still be two stages.  just the first one ends up being bypass-able<a href="#t13:32" class="time">13:32</a>
nottingjeremy: single-image :P<a href="#t13:33" class="time">13:33</a>
jeremy(which matches what we do if you boot off the dvd; just in that case, we use the dvd repodata and won't need to do the asking like we will here)<a href="#t13:33" class="time">13:33</a>
f13jeremy: think that will be ready by Test1?<a href="#t13:33" class="time">13:33</a>
f13or at least testable?<a href="#t13:33" class="time">13:33</a>
jeremyf13: umm... maybe? :)  it's not that much work.<a href="#t13:34" class="time">13:34</a>
f13jeremy: well, it matters for what we'll deliver for Test1.  If we can just deliver a fully installable Fedora spin+tree and just an Everything repo, great!<a href="#t13:35" class="time">13:35</a>
jeremyf13: how about I state it as "if we want it, we can probably get it done for then".  although somehow, I have this sinking feeling that I'm signing myself up to do something ;-)<a href="#t13:36" class="time">13:36</a>
f13sounds good to me.<a href="#t13:37" class="time">13:37</a>
* rdieter needs to duck out to another meeting in a few minutes...<a href="#t13:37" class="time">13:37</a>
f13so, lets recap a bit on this topic.<a href="#t13:37" class="time">13:37</a>
f13Rel-eng will deliver a Fedora spin which is iso(s) + exploaded installable tree.  We'll deliver a set of Fedora Live images for i386/x86_64 that are installable.  We'll deliver KDE Live images that are installable, and we'll deliver an Everything repo of packages that we can hopefully use with other install media to "make" it installable.<a href="#t13:39" class="time">13:39</a>
f13am I missing anything?  wwoods these are the things we'd be on the hook to QA as well<a href="#t13:39" class="time">13:39</a>
wwoodsright<a href="#t13:39" class="time">13:39</a>
wwoodsQA procedure for the Live spins is not well-defined at this point<a href="#t13:40" class="time">13:40</a>
wwoodsgotta add more of that stuff to the test matrices and whatnot<a href="#t13:40" class="time">13:40</a>
wwoodsbut yeah, that's fine<a href="#t13:42" class="time">13:42</a>
f13jeremy: notting: rdieter_away: jwb: am I missing anything that you can think of?<a href="#t13:43" class="time">13:43</a>
jeremynot I<a href="#t13:43" class="time">13:43</a>
nottingunless you want to spin off some of the live images to other groups entirely<a href="#t13:44" class="time">13:44</a>
nottingalso, doesn't actually say the driving idea behind what goes in the 'fedora' spin. dunno if that's your problem<a href="#t13:44" class="time">13:44</a>
f13notting: it's still gotta come to us if we want to gpgsign the sha1sum file.<a href="#t13:44" class="time">13:44</a>
f13notting: hee, content definition.  Not My Problem?  (:<a href="#t13:45" class="time">13:45</a>
wwoodsIt'd be cool if (e.g.)  KDE SIG was mostly responsible for the spinning / QA / etc on the KDE spin<a href="#t13:45" class="time">13:45</a>
f13I'd be cool with that as well, adn they're in charge of the content definition for that spin mostly<a href="#t13:46" class="time">13:46</a>
warrenIs rdieter_away in this rel-eng group?<a href="#t13:46" class="time">13:46</a>
f13yes<a href="#t13:46" class="time">13:46</a>
jeremywwoods: that's one reason I want to have us _stop_ composing on local workstations<a href="#t13:46" class="time">13:46</a>
nottingfor practical purposes, spins we release at <release> date are much easier to generate by someone in the rel-eng group with access to .fp.o<a href="#t13:46" class="time">13:46</a>
jeremynotting: on .fp.o, yes.  but not reducto.boston.redhat.com :-)<a href="#t13:47" class="time">13:47</a>
f13heh<a href="#t13:48" class="time">13:48</a>
wwoodswell, most of the tools and everything are public, aren't they? maybe we just need better process docs?<a href="#t13:48" class="time">13:48</a>
f13I'll still wind up doing some composing there for development of hte tools.<a href="#t13:48" class="time">13:48</a>
f13but yeah, "official" spins should happen in fp.o space.<a href="#t13:48" class="time">13:48</a>
f13just so that the output is somewhat publically available.<a href="#t13:49" class="time">13:49</a>
f13although -ESPACE right now.<a href="#t13:49" class="time">13:49</a>
* mmcgrath has submitted a ticket about our current issues btw<a href="#t13:49" class="time">13:49</a>
f13poelcat: Decision: Rel-eng will deliver a Fedora spin which is iso(s) + exploaded installable tree.  We'll deliver a set of Fedora Live images for i386/x86_64 that are installable.  We'll deliver KDE Live images that are installable, and we'll deliver an Everything repo of packages that we can hopefully use with other install media to "make" it installable.<a href="#t13:49" class="time">13:49</a>
f13poelcat: Various SIGs may be responsible for content definition for these various deliverables, as well as pre-generation and QA of them.  rel-eng will just be the final producers and signers of the content.<a href="#t13:50" class="time">13:50</a>
f13poelcat: at least one week before Test1 freeze rel-eng will be looking closer and attempting to fix up issues such as broken deps, broken upgrade paths, conflicts, and other such tree issues.  Compose tools should be somewhat stabalized before this point so that pre-freeze trees can be generated and reviewed before the mad freeze rush.<a href="#t13:52" class="time">13:52</a>
Idea.png f13 changed the topic of #fedora-meeting to: Fedora Release Engineering Meeting - Open Discussion<a href="#t13:52" class="time">13:52</a>
f13FYI I created <a href="http://fedoraproject.org/wiki/ReleaseEngineering/SOP">http://fedoraproject.org/wiki/ReleaseEngineering/SOP</a><a href="#t13:52" class="time">13:52</a>
f13and an example SOP (RemovingPackages).  I'm thinking about making a Template out of the RemovingPackages page so that we can easily generate pages for other tasks.<a href="#t13:53" class="time">13:53</a>
f13any other topics anybody would like to discuss?<a href="#t13:54" class="time">13:54</a>
f13<crickets><a href="#t13:55" class="time">13:55</a>
nottingdo we need to make sure to get anything update-related on the fesco schedule this week?<a href="#t13:55" class="time">13:55</a>
wwoodspackage update stuff? like bodhi-related thingers?<a href="#t13:56" class="time">13:56</a>
nottingwhether it's automatically doing the buildroots from dist-fc7-updates-candidate, or auto-pusing after X days, or...<a href="#t13:56" class="time">13:56</a>
f13notting: buildroot stuff is held up on bodhi being able to prevent things from going out that were built with things that aren't out yet<a href="#t13:57" class="time">13:57</a>
wwoodsauto-dep-checking tools for new updates, yeah<a href="#t13:57" class="time">13:57</a>
f13notting: auto-pushing we haven't really discussed how that would work<a href="#t13:57" class="time">13:57</a>
wwoodsis there a (prototype) tool yet?<a href="#t13:57" class="time">13:57</a>
f13and the entire dep-checking thing is still up in teh air looking for a way that doesn't completely suck or take a day to complete.<a href="#t13:57" class="time">13:57</a>
f13wwoods: I don't think so<a href="#t13:57" class="time">13:57</a>
nottingf13: mbonnet: what do we do static repos for?<a href="#t13:57" class="time">13:57</a>
wwoodsI've been hacking around with some stupid junk for that but I wanted to get some folks who knew what they were doing involved<a href="#t13:58" class="time">13:58</a>
nottingis it just a list, or is it 'tags that have property X'?<a href="#t13:58" class="time">13:58</a>
f13notting: a list<a href="#t13:58" class="time">13:58</a>
nottingok<a href="#t13:58" class="time">13:58</a>
mbonnetnotting: just a list, synced by a cronjob at the moment<a href="#t13:58" class="time">13:58</a>
lmackenso I've got multi-build support, among a *ton* of other enhancements/fixes in my local dev branch.. the only thing left with that is to tweak the new update form<a href="#t14:01" class="time">14:01</a>
lmackeni've been working with upstream of the TGExpandingWidget to fix some problems.. but I'm probably just going to write my own, like the old system has<a href="#t14:02" class="time">14:02</a>
lmackenbut thanks to insane amounts of work for school, I wasn't able to put any cycles into bodhi over the weekend.. I'll hopefully have all of that stuff out the door this week<a href="#t14:02" class="time">14:02</a>
lmacken(for shits and giggles, I started writing a forward-chaning rule-engine expert system (using the rete algorithm) to try and solve deps/closure)<a href="#t14:05" class="time">14:05</a>
wwoodswhich is awesome<a href="#t14:05" class="time">14:05</a>
wwoodsheh<a href="#t14:05" class="time">14:05</a>
f13ok, meeting over.<a href="#t14:05" class="time">14:05</a>
lmacken(mainly for my AI class.. but if it works.. hey, why not)<a href="#t14:05" class="time">14:05</a>
mclasenlmacken: fear the thinking repository<a href="#t14:06" class="time">14:06</a>
lmackenmclasen: :)<a href="#t14:06" class="time">14:06</a>
Idea.png f13 changed the topic of #fedora-meeting to: <a href="http://fedoraproject.org/wiki/Communicate/FedoraMeetingChannel">http://fedoraproject.org/wiki/Communicate/FedoraMeetingChannel</a> -- Meetings often get logged -- see schedule in the wiki for next meeting<a href="#t14:06" class="time">14:06</a>

Generated by irclog2html.py 2.3 by Marius Gedminas - find it at mg.pov.lt!